Community detection is a fundamental task in social network analysis. Online social networks have dramatically increased the volume and speed of interactions among users, enabling advanced analysis of these dynamics. Despite a growing interest in tracking the evolution of groups of users in real-world social networks, most community detection efforts focus on communities within static networks. Here, we describe a framework for tracking communities over time in a dynamic network, where a series of significant events is identified for each community. To this end, a modularity-based strategy is proposed to effectively detect and track dynamic communities. The potential of our framework is shown by conducting extensive experiments on synthetic networks containing embedded events. Results indicate that our framework outperforms other state-of-the-art methods. In addition, we briefly explore how the proposed approach can identify dynamic communities in a Twitter network composed of more than 60,000 users, which posted over 5 million tweets throughout 2020. The proposed framework can be applied to different social network and provides a valuable tool to understand the evolution of communities in dynamic social networks.

翻译：社区检测是社交网络分析中的基础任务。在线社交网络极大地提升了用户间互动的数据量与交互速度，使得对这些动态过程的深入分析成为可能。尽管追踪真实社交网络中用户群体演化的研究兴趣日益增长，但大多数社区检测工作仍聚焦于静态网络中的社区。本文提出了一种动态网络中随时间追踪社区的框架，该框架可为每个社区识别一系列关键事件。为此，我们提出了一种基于模块度的策略，用以有效检测并追踪动态社区。通过在包含预设事件的合成网络上进行广泛实验，验证了该框架的潜力。结果表明，我们的框架优于其他现有先进方法。此外，我们初步探讨了该方法如何在由超过6万名用户（2020年期间发布逾500万条推文）构成的Twitter网络中识别动态社区。该框架可适用于不同类型的社交网络，为理解动态社交网络中社区的演化提供了有效工具。